二维肋片换热FORTRAN 源程序

内聚力的循环本构开发，新的UEL单元，比较复杂，很难学会的(Cyclic constitutive development of cohesion)

ABAQUS中用户材料子程序UMAT的开发主要解决两方面的问题：本构模型的建立和积分算法的选择。 本文主要研究非线性材料的UMAT实现方法，并重点研究其迭代算法部分，目前，用户材料子程序UMAT的迭代算法主要是常刚度法，常刚度法的优点在于算法原理较简单，程序编写较方便，缺点是当遇到复杂非线性材料时，其迭代次数较多，收敛速度也较慢，在这个情况下，本文采取的是一种迭代次数较少且收敛速度较快的切线刚度法，具体就是采用FORTRAN语言编制了基于Von-Mises模型的接口程序,并采用切线刚度算法，通过与ABAQUS自带本构关系计算的结果相比较，验证其正确性。(This paper studies the user subroutine UMAT of ABAQUS development methods, the use of FORTRAN language isotropic hardening material model of the interface program, studied the effects of such material is extremely elastic-plastic constitutive relation method. This article UMAT tightly around the secondary development of technology, the first principle of its interface detail, and then for the theory of nonlinear finite element incremental stiffness of the regular tangent stiffness method and the theory of algorithms to do an in-depth analysis of deduced a regular tangent stiffness and rigidity of the law of the specific expression of algorithm theory, and then the preparation of the two algorithms, respectively, of the UMAT program, and finally the establishment of a specific model checking, bringing with ABAQUS elasto-plastic constitutive relation of the calculated results compared to verify the correctness of the two.)

Umat friction M-C criterior

creep 子程序编程规范示例，在ABABQUS用户使用手册有清晰说明与注释(Creep subroutine programming specification example)

说明：  uhard subroutine code in fortran for abaqus

将autocad软件中的线和面导入到ansys模型中，进行建模和处理(transform line and area from autocad to ansys)

说明：  CUDA Fortran 高效编程实践_科学家和工程师特供 2007 年以来，以 nVidia GPU 为代表的加速器并行计算风起云涌，带有加速器的超级计 算机在 TOP500 中的份额逐年增加，支持加速器的主流应用软件也呈爆炸式增长，研究加速 器计算的技术人员数以百万计，世界范围内的大学、研究机构竞相开设相关课程。 目前流行的 GPU 通用编程语言是 CUDA C 和 OpenCL. 它们均是 C/C++语言的扩展，因 此可以方便地将 C/C++代码移植到 GPU 上。但对于科学与工程计算中的重要编程语言 Fortran，无法直接地改写为 CUDA C 或 OpenCL。很多久经考验的应用程序都是用 Fortran 开发，如果完全改写，工作量巨大，而且有不可估量的稳定性风险。 为使 Fortran 应用能够使用 GPU 加速，The Portland Group 设计了 CUDA Fortran 语言， 并在自家的 PGI 编译器中支持。气象、理论物理等领域的应用经过简单的改造，就能够利 用 GPU 的强大计算能力。(CUDA FORTRAN efficient programming practice Since 2007, with NVIDIA GPU as the representative of accelerator parallel computing, the share of supercomputers with accelerators in top 500 has increased year by year, and the mainstream application software supporting accelerators has also grown explosively. There are millions of technicians studying accelerator computing, and universities and research institutions around the world are competing to offer relevant courses. To enable FORTRAN applications to use GPU acceleration, the Portland group designed CUDA FORTRAN language, which is supported in its own PGI compiler. The application of meteorology, theoretical physics and other fields can make use of the powerful computing power of GPU after simple transformation.)

说明：  LsDyna二次开发流程介绍及示例。umat42各向同性材料。(Introduction and example of secondary development process of LSDYNA)

用TDMA方法制作的程序，一般用于求解难度较大的换热器问题。(Program produced with the TDMA method, generally used for solving the difficult problem of the heat exchanger.)

a simple ct & pt model in pscad.